The blind recognition of channel coding is playing more and more important roles in smart communications. The paper took rate below 1/2 binary cyclic codes to research, we established a model using counter of continuous bits 0. At the basic of studying the features of binary cyclic codes, when the condition about the rate is satisfied, we made a statistics of the continuous bits 0 in a long sequence of code words to get the length of code word n and the length of information k. In the condition of higher BER simulations, we improved the method to make it suitable for estimating the parameter.
Performance Analysis of Deep Learning based on Recurrent Neural Networks for Channel Coding
